2008-09-01, 01:03 AM
Normal distribution is ironically an approximation of xhko.
Binomial distribution is actually not relevant to this purpose. It's useful when there are two possibilities (e.g. flipping a coin or passing a scroll), and spits out all the possibilities given the number of trials and probability of each outcome (heads/tails, pass/fail). But when you've got a damage range that isn't 1-2 or 9-10 or 32767-32768, it becomes useless.
What would really be useful for calculating this would be a function that spits out the distribution of a sum of dice rolls, given the number of sides on the dice and the number of dice rolled.
I'm currently trying to work something out with nested arrays in Mathematica to see if it could spit out some numbers for me, but even if I manage to get it to work (and I'm doubtful) my computer might not appreciate doing all that number bashing.
Binomial distribution is actually not relevant to this purpose. It's useful when there are two possibilities (e.g. flipping a coin or passing a scroll), and spits out all the possibilities given the number of trials and probability of each outcome (heads/tails, pass/fail). But when you've got a damage range that isn't 1-2 or 9-10 or 32767-32768, it becomes useless.
What would really be useful for calculating this would be a function that spits out the distribution of a sum of dice rolls, given the number of sides on the dice and the number of dice rolled.
I'm currently trying to work something out with nested arrays in Mathematica to see if it could spit out some numbers for me, but even if I manage to get it to work (and I'm doubtful) my computer might not appreciate doing all that number bashing.